huggingface/transformers

Pull Requests Commits

Fix cache type in Idefics2 (#30729)

zucchini-nlp committed 1 year ago

Verified c02d302e

Fix OWLv2 Doc (#30794)

jla524 committed 1 year ago

Verified 449894d2

CI: update to ROCm 6.0.2 and test MI300 (#30266)

fxmarty committed 1 year ago

Verified 37bba2a3

skip low_cpu_mem_usage tests (#30782)

SunMarc committed 1 year ago

Verified 539ed75d

Deprecate models script (#30184)

amyeroberts committed 1 year ago

Verified 0f8fefd4

Save other CI jobs' result (torch/tf pipeline, example, deepspeed etc) (#30699)

ydshieh committed 1 year ago

Verified 82c1625e

Generate: assistant should be greedy in assisted decoding (#30778)

gante committed 1 year ago

Verified 2e27291c

Port IDEFICS to tensorflow (#26870)

a8nova committed 1 year ago

Verified 94306352

Generate: remove near-duplicate sample/greedy copy (#30773)

gante committed 1 year ago

Verified de2f7221

[Object detection pipeline] Lower threshold (#30710)

NielsRogge committed 1 year ago

Verified ce87dca1

enable Pipeline to get device from model (#30534)

faaany committed 1 year ago

Verified 69d9bca5

Qwen: incorrect setup flag (#30776)

gante committed 1 year ago

Verified f4dc26d4

Generation / FIX: Fix multi-device generation (#30746)

younesbelkada committed 1 year ago

Verified f823fec5

Llama: fix custom 4D masks, v2 (#30348)

poedator committed 1 year ago

Verified a0779b9e

[GroundingDino] Adding ms_deform_attn kernels (#30768)

EduardoPach committed 1 year ago

Verified 453893ed

Support for Falcon2-11B (#30771)

Nilabhra committed 1 year ago

Verified e52741f6

Blip dynamic input resolution (#30722)

zafstojano committed 1 year ago

Verified f63d8222

Workflow: Replace `actions/post-slack` with centrally defined workflow (#30737)

younesbelkada committed 1 year ago

Verified a4e530e3

[awq] replace scale when we have GELU (#30074)

SunMarc committed 1 year ago

Verified de6e0db1

hqq - fix weight check in check_quantized_param (#30748)

mobicham committed 1 year ago

Verified e0c3cee1

[docs] Update link in es/pipeline_webserver.md (#30745)

aaronjimv committed 1 year ago

Verified 8ce4fefc

PEFT / Trainer: Make use of `model.active_adapters()` instead of deprecated `model.active_adapter` whenever possible (#30738)

younesbelkada committed 1 year ago

Verified 2d1602ae

mlp_only_layers is more flexible than decoder_sparse_step (#30552)

eigen2017 committed 1 year ago

Verified 1c52cb7b

Update llama3.md, fix typo (#30739)

mimbres committed 1 year ago

Verified 73fcfb28

[docs] Update es/pipeline_tutorial.md (#30684)

aaronjimv committed 1 year ago

Verified 47735f5f

Update CodeLlama references (#30218)

osanseviero committed 1 year ago

Verified c99d88e5

Generate: consistently handle special tokens as tensors (#30624)

gante committed 1 year ago

Verified 7130a22d

KV cache is no longer a model attribute (#30730)

zucchini-nlp committed 1 year ago

Verified 5413b898

Fix image post-processing for OWLv2 (#30686)

jla524 committed 1 year ago

Verified 218f4413

Generate: add `min_p` sampling (#30639)

gante committed 1 year ago

Verified df53c6e5

Newer Older